european portuguese possessives and the structure of dp

Cuadernos de Lingüística del I. U. I. Ortega y Gasset 2007, vol. 14, pp. 27-50 27 © 2007 IUIOG, Irene Gil, Héctor Fernández, Raquel González

European Portuguese possessives and the structure of DP*

ANA MARIA BRITO Abstract

The main aim of this paper is to analyse, in a comparative perspective, the category of possessives in European Portuguese possessives as for their X and XP nature and to relate their internal to their external syntax within the DP structure, mainly centred on the possibilities of co-ordination and focalisation of possessives. It will be proposed that both pre-nominal and post-nominal possessives can be focalised, that the nature of contrastive focus that affects pre-nominal possessives does not argue in favour of their X nature and that the focalisation of pre-nominal and post-nominal possessives does not justify any functional category related with informational structure in the DP. 1. Introduction

In recent years there has been a significant research on possessives in European Portuguese (EP). One of the discussed matters is the weak or strong nature of possessives. One perspective considers that pre-nominal possessives are weak and heads (Castro and Costa 2002, 2003; Castro 2005); another perspective, although considering that pre-nominal possessives are weak and post-nominal possessives are strong, propose that both are maximal projections, but it is also assumed that a reanalysis process is occurring, the head nature dominating the maximal projection nature of pre-nominal possessives (Miguel 2002 a, b, c, 2004).

The main aim of this paper is to analyse, in a comparative perspective, the category of possessives in European Portuguese possessives as for their X and XP nature and to relate their internal to their external syntax, mainly centred on the possibilities of coordination and focalisation of possessives.

The text is organised in the following way: in section 2. I make a brief presentation of some of the diagnosis for the distinction between weak and strong person pronouns and possessives. The analysis of possessives presented in sections 3 to 5 allows us to indicate that there are grammars in competition in EP with respect to pre-nominal possessives. The syntactic analysis presented in 6 also includes the study of post-nominal possessives. The paper ends with a synopsis of possessives in EP (7) and with the conclusions (8). 2. Background

2.1. Weak and strong person pronouns

Grammarians and linguists have very often been sensitive to the different external syntax of person pronouns, but only since Postal (1969), Raposo (1973), Kayne (1975), Cardinaletti and Starke (1994 / 1999), Cardinaletti (1999) and more recently Déchaine and Wiltschko (2003), have we known that they differ also in their internal syntax. Kayne has shown that person pronouns are not uniform and that there is a distinction between strong pronouns and clitics: strong pronouns can be contrastively stressed or used in isolation, coordinated and modified by

* I thank Gabriela Matos and Petra Sleeman for their comments to a previous version of this paper; all errors are

mine.

Ana Maria Brito

28

certain adverbs, in contrast with clitics, as Italian demonstrates in (1) to (4) (cf. Cardinaletti 1999: 34-5): 1

(1) a. Maria conosce LUI, non voi. Mary knows HIM not you b. * Maria LO conosce, non voi. Mary IT[HUM/-HUM] knows not you

(2) a. Chi conosce, Maria? Lui. Who you.know Mary? Him. b. * Chi conosce, Maria? Lo. Who you.know Mary? It[HUM/-HUM]

(3) a. Maria conosce lui e voi. Mary knows him and you b. * Maria lo e vi conosce. Mary it[hum/-hum] and you know

(4) a. Maria conosce solo lui. Mary knows only you b. * Maria lo conosce solo. Mary it[HUM/-HUM] knows only

Cardinaletti and Starke (1994/1999) proposed that the notion of a deficient pronoun may include two types, weak person pronouns, that are considered to be maximal projections, and clitics, that are heads. The notion of weak pronoun was in the first place proposed for Germanic languages, which have pronouns, different from clitics, but with a syntax different from strong pronouns and full noun phrases (Cardinaletti 1999: 33, 60):

(5) a. It strikes me as implausible. b. *What strikes you as implausible? It. c. *It, I think is implausible. d. *IT is implausible, not that one.

In Portuguese, we also have three classes of person pronouns: those that are neither clitic nor strong: eu ‘I’, tu ‘you’, nós ‘we’, vós ‘you plural’; there are clitics: me ‘me’, te ‘you DO and IO’, se ‘3rd person reflex’, o ‘it masc. sg.’, a ‘it fem. sg.’, os ‘it masc. pl.’, as ‘it fem. pl.’; and there are strong person pronouns: (co)migo ‘with me’, (con)tigo ‘with you’, ele ‘he’, eles ‘they masc. pl.’, ela ‘she’, elas ‘they fem. pl.’.

The following examples illustrate the behaviour of the three series of pronouns vis à vis stress, isolated use, coordination and modification; in (6) non-clitic and non-strong pronouns, in (7) clitics and in (8) strong pronouns:

(6) a. EU vou sair, não a Maria. ‘I am leaving, not Mary.’ b. Quem está aí? EU. ‘Who is there? I.’ c. Eu e tu vamos sair. ‘I and you are leaving.’ d. Só eu vou sair. ‘Only I am leaving.’

1 For some examples only the glosses will be presented, because the translations are irrelevant; the capital letters in

the examples and glosses indicate heavy stress.

European Portuguese possessives and the structure of DP

29

(7) a. *A Maria conhece-A. ‘Mary knows it.FEM.SG[HUM/-HUM].’ b. *Quem está aí? A. ‘Who is there? It.FEM.SG.’ c. *A Maria conhece-a e o. ‘Mary knows it.FEM.SG and it.MASC.SG.’ d. *A Maria conhece só a. ‘Mary knows only it.FEM.SG.’

(8) a. ELA saiu. SHE left b. Quem está aí? Ela. ‘Who is there? (It is) her.’ c. Ele e a Maria saíram. ‘He and Mary left.’ d. Só ele veio. ‘Only he came.’

In a Principles and Parameters framework, the distinction mentioned above has been developed in different ways: one proposal, made by Cardinaletti and Starke at the end of their classical paper, is that for the three classes of person pronouns there are corresponding configurations: strong pronouns appear in CP, weak pronouns in �P and clitics in IP, that is, under this view, morphological deficient elements realise less syntactic structure (p. 214).

Another view, related to X’ Theory and suggested by Cardinaletti (1999), is that clitics are heads (X), while weak and strong pronouns are maximal projections, XP, projected at different positions in the syntactic structure.

More recently, the discussion on the internal structure of person pronouns was developed by Déchaine and Wiltschko (2003). Starting with the data presented in Postal (1969), according to which some person pronouns behave as determiners, the authors propose a distinction between pro-DPs, pro-�P and pro-NP. The first and second English person pronouns we and you can function as Ds, as in (9a) and (9b) and then they instantiate pro-DP; third person pronouns cannot function as Ds, as in (9c), and they instantiate pro-�P; and one has a distribution similar to a N and is analysed as pro-NP. (10) represents the internal structure of these three person pronouns:

(9) a. we linguists; us linguists. b. you linguists. c. *they linguists; *them linguists. d. the one; some one, the real one.

(10) a. [DP we [�P � [NP 0 / linguists ]]] b. [�P � [� they ]] c. [NP one]

We see that, in Cardinaletti’s terms, only clitics are X and both weak and strong pronouns are XP, although differing in their external syntax. For Déchaine and Wiltschko pronouns differ in their internal structure and the different categorial status of the three classes determines their distribution.

Ana Maria Brito

30

2.2. Weak and strong possessives

Some languages exhibit a differentiation between two series of possessives: one series of phonologically weak possessives in pre-nominal position and another series of phonologically strong possessives in post-nominal position and predicative contexts. It is the case of Spanish, exemplified in (11):

(11) a. su casa his/her house b. la casa suya the house her

Other languages have similar forms for possessives, although used in different contexts: pre-nominal, post-nominal and predicative. This is the case of Italian and Portuguese, where there is no apparent evidence for postulating different classes of possessives. Nevertheless, Cardinaletti (1998) has proposed that pre-nominal and post-nominal Italian possessives, illustrated in la sua casa and la casa sua, though exhibiting the same superficial form, are different constituents. In pre-nominal position, possessives cannot be modified by exclusion adverbs (12a) nor coordinated (12b) and may have a human and non-human reading (12c):

(12) a. *la solo sua casa the only his/her house b. *la sua e di Maria casa the his and of Mary house c. Il suo coperchio è molto pratico. the its/his/her[human and non human] lid is very pratical

In post-nominal position, possessives may be modified by exclusion adverbs (13a), may be coordinated (13b) and must have a human reading (13c):

(13) a. la casa solo sua the house only his/her b. la casa sua e tua the house his/her and yours c. Il coperchio suo è molto prático. the lid his/her[human] is very pratical ‘his/her lid is very pratical’

Based on these examples, the author proposes that pre-nominal possessives are weak forms and post-nominal ones are strong forms. Cardinaletti also uses differences in focus, but, for the moment, I will not present her data. Relating these different behaviours with the internal and the external syntax of possessives, the author proposes that pre-nominal and post-nominal possessives are not transformationally related; the pre-nominal possessives are XP, generated in the Spec of a functional category in the DP (AgrSN), while the post-nominal possessives are XP generated at the right side of the DP (p. 22). Cardinalleti also notes that in some Italian dialects there are possessive forms that exhibit a weak form from the phonological point of view and that allow doubling, as in (14), which makes possible for her to propose that they are clitics, presumably adjoined to D:

(14) so pare de Toni (Padova) his father of Toni

This shows that Italian has a tripartition in possessives (clitic, weak and strong) and not only two types of forms. As a first comment to this analysis, it is important to note that some of the criteria proposed by Cardinaletti for the distinction between weak and strong possessives (exclusion adverbs, coordination) are the same used by Kayne for the distinction between clitics


31

and non-clitics, and then what we would expect from her argumentation would be that some Italian possessives tend to cliticize. We see that the terminology weak / strong is misleading and must be replaced by more formal approaches. In the following sections we will study possessives in EP. 3. Pre and post-nominal possessives in EP: a first approach

Pre-nominal possessives are normally used in EP with definite articles, as in (15a); post-nominal possessives are used with indefinites (15b), as well as with numerals, null determiner, interrogative and exclamative wh words (cf. Brito 1984, 2001; Castro 2005):

(15) a. o meu livro the my book ‘my book ‘ b. um livro meu a book my ‘a book of mine’

If the examples shown in (15) constituted the totality of the relevant data it would be possible to make correlations between definiteness and pre-nominal possessives and between indefiniteness and post-nominal possessives. In fact, for many speakers of EP, only the examples in (15) are good. But for others the pre-nominal possessive is accepted with indefinites, as in (16) (Brito 2001: 560, 2003: 509; Miguel 2002a: 294):

(16) Uma minha amiga saiu. a/one my friend left ‘One of my friends left.’

These behaviours show that there is variation in the syntax of pre-nominal possessives in EP: one that seems more “conservative” and the “standard” one (Miguel 2002a, b, c, 2004). The so-called “conservative” grammar exhibit one of the orders that we can find in medieval texts, as in the example (17) (apud Mattos e Silva 1989: 183):

(17) h�� nosso menino a our child

It may also be that the two positions are associated with semantic differences. In fact, Miguel (2002c: 236-7) shows that the availability of pre-nominal possessives with indefinite articles is limited to a specific reading (18a) and it is excluded from negative and opaque contexts; see (18b):

(18) a. Um meu amigo trouxe-me uma prenda de Paris. a my friend brought-me a gift from Paris ‘A friend of mine brought me a gift from Paris.’ b. *? Um meu amigo não diria isso. a my friend wouldn’t say that ‘A friend of mine wouldn’t say that.’

While the post-nominal possessive accepts the two readings, as illustrated in (19) (idem):

(19) a. Um amigo meu trouxe-me uma prenda de Paris. a friend my brought-me a gift from Paris ‘A friend of mine brought me a gift from Paris.’ b. Um amigo meu não diria isso. a friend my wouldn’t say that ‘A friend of mine wouldn’t say that’

Ana Maria Brito

32

The presentation made until now shows that EP has at least two grammars of pre-nominal possessives. The non-dominant grammar2 allows pre-nominal possessives with indefinites (although limited to a specific reading), the dominant grammar only accepts pre-nominal possessives with definites and post-nominal ones with indefinites.3 4. Some arguments presented in the literature for the distinction between weak and

strong possessives in European Portuguese

The examples in (15a) and (15b) show that, similarly to Italian, Modern Portuguese presents the same form for possessives in pre-nominal and post-nominal positions. Apparently we have the same forms in the two positions. But, based on the analysis of Italian by Cardinaletti (1998), Castro and Costa (2002, 2003), Castro (2005) and Miguel (2002a, b, c, 2004) have shown that, despite their similar form, pre-nominal and post-nominal possessives in Portuguese differ in several aspects. Let us present some of the relevant criteria.

4.1. Modification, human reading and reduction

I) Modification by exclusion adverbs like só, apenas ‘only, just’ (Castro and Costa 2002: 104):

(20) a. *O só meu problema é que não percebo nada disto. the only my problem ‘is that I do not understand it.’ b. Um problema só meu é que não percebo nada disto. ‘A problem only mine is that I do not understand it.’

The contrast shows that a pre-nominal possessive cannot be modified by exclusion adverbs, unlike what happens to a post-nominal possessive.

II) The [+ / - human] reading (Castro and Costa 2002: 104):

(21) a. A suai tampa é muito prática. (i = Maria/frying pan) its/her lid is very practical b. Encontrei uma tampa suai. (i = Maria/*frying pan) I found a lid hers ‘I found a lid of hers’

The examples show that a pre-nominal possessive may refer to a human or non-human entity, but the post-nominal possessive must refer only to human entities.

III) Reduction (Castro and Costa 2002: 104; Miguel 2002a: 296; 2004: 405):

(22) a. o meu livro the my (non reduced) book ‘my book’

2 There is no sociolinguistic study of these grammars in competition, so their classification is provisional. The non-

dominant grammar mainly corresponds to the grammar of middle-aged speakers, in a written and formal discourse, with special incidence from North and Centre of Portugal. The dominant grammar mainly corresponds to the grammar of young speakers, in an oral and informal discourse, with special incidence from southern towns.

3 The web searches done by Castro (2005) show that “combined with the definite article, simple possessives tend to be prenominal whereas with the indefinite article they are mainly postnominal.” (p. 56) In the Subcorpus Oral Espontâneo (SCOE), the author found the following values: for 151 possessivised noun phrases, 137 were prenominal, 123 of them headed by definite article, 6 with demonstrative and 8 determinerless (pp. 56-7). Miguel (2004: 403) explicitly affirms that the grammar she classifies as “standard” is not yet the dominant dialect, but the numbers presented above by Castro are very representative, meaning that her grammar of possessives is becoming the dominant one.


33

b. o me livro the my (reduced) book ‘my book’ c. um livro meu a book my (non reduced) ‘a book of mine’ d. *um livro me a book my (reduced) ‘a book of mine’

The examples show that a pre-nominal possessive may be subject to a reduction process (22b), but this process is impossible with a post-nominal possessive (22d). The reduction characterises some southern and oral dialects of European Portuguese. Based on these criteria, Castro and Costa (2002, 2003) and Castro (2004) proposed that pre-nominal possessives are weak and are heads (X); and post-nominal possessives are strong and maximal projections (XP). Miguel (2002 a, b, c, 2004) makes a different approach: only the reduced forms of possessives, as in (22b), are considered to be heads; pre and post-nominal possessives are assumed to be, respectively, weak and strong, but the two series are classified as maximal projections. In order to describe the “standard” grammar she proposes that a reanalysis process is occurring, the head nature dominating the maximal projection nature of pre-nominal possessives4. Then, Miguel’s analysis for Portuguese is very similar to the one of Cardinaletti for Italian, with the proposal of a tripartition of possessives and not of two classes of possessives.

We will come back to this matter later. Meanwhile let us analyse some more problematic criteria for the distinction between weak and strong forms: coordination (3.2) and focus (3.3).

4.2. Coordination associated to focus

Cardinaletti and Starke (1994 / 1999: 145, 149), discussing the nature of personal pronouns, propose that deficient pronouns (in the examples, il in French and esse in Italian), are incompatible with coordination:

(23) *Il et celui de Jean sont beaux. he and that one of John are beautiful

(24) *Esse e quelle accanto sono troppo alte. these and those acanthus are too high

And Cardinaletti (1998: 19), on Italian possessives, shows that pre-nominal possessives cannot be coordinated, as in (25), allowing her to consider that they are weak possessives:

(25) *la sua e tua casa the his and your house

Let us see how Portuguese behaves as for this phenomenon. The following example is a very interesting case of coordination of pre-nominal possessives, because it is a clear instance of reference to a single entity:

(26) “O seu (e meu) amigo Pacheco Pereira (…) é capaz de dizer (…) que ele prefere tomar medidas sociais e culturais.” (Eduardo Prado Coelho, Público, 23th October 2006, p. 6)

the his and my friend P.P. is able to say that (...) ‘His and my friend P.P. is able to say that (...)’

4 Miguel (2004) does not clarify if this head is an A or a D, but in other work (2002b) her proposal is that

possessives in EP are “personal adjectives”.

Ana Maria Brito

34

Other examples of coordination of possessives in pre-nominal position seem also possible, both of non-correlative and correlative coordination (27a, b), although the examples are always better with coordination of post-nominal possessives (28a, b) (Miguel 2002a: 294, Brito 2003: 506):

(27) a. ? O meu e TEU artigo está aqui. the my and YOUR paper is here ‘My and your paper is here.’ b. ? O não só meu mas também TEU artigo está aqui. the not only my but also YOUR paper is here ‘My and your paper is here.’

(28) a. Um artigo meu e TEU foi publicado na revista. a paper mine and YOURS was published in.the journal ‘A paper of mine and yours was published in the journal.’ b. Um artigo não só meu mas também TEU foi publicado na revista. a paper not only mine but also YOURS was published in.the journal ‘A paper of mine and yours was published in the journal.’

Castro and Costa (2002: 104) proposed that coordination of possessives is possible if the second member of the coordination is focalised, a proposition that seems to be on the right track, but that was not fully explored. In fact, the coordination of pre-nominal possessives raises a theoretical problem for their analysis, because these possessives are behaving as strong forms. Moreover, since Kayne (1994) it is generally accepted that there is no coordination of heads, only of XP.5 So, what araises from this possibility is that, in the context of coordination, pre-nominal possessives are XP and that when the speakers of the dominant grammar of possessives accept these data they are using, in fact, the non-dominant and more formal grammar, a conclusion that seems plausible due to the idea that grammars change gradually. In these circumstances, the simplified syntactic structure of the constituent containing the coordination of possessives in (26) is presented in (29) (see Kayne (1994) on coordination and section 6. for AgrP in DP):

(29) DP ei D AgrP | ei o ConjP Agr’ ru | XP Conj’ Agr | | ru seu Conj XP NP | | | e meu amigo ...

4.3. Focus on possessives

4.3.1. Some general remarks As the notions of informational and contrastive focus (Kiss 1996) will be crucial in our analysis, it is important to distinguish clearly between them. The informational focus is the new information introduced in the discourse and contrastive focus is the one where the set of

5 Gabriela Matos pointed out to me that there is coordination at the morphologic level, as in Lenta mas

corajosamente, a Maria leu aquele texto denso ‘Slow but courageously, Mary read that dense text’, which may suggest that there is a coordination of heads.


35

alternatives that is required for the interpretation of a sentence contains at least one member. Let us take the examples (30a) and (30b):

(30) a. (John called) Mary. b. John called MARY. (with focus on Mary) c. Who did John call?

In (30a) the informational focus is Mary, as an answer to (30c). In (30b), with a stress on Mary, the semantic interpretation requires a contrast set of propositions of the form ‘John called x, x being a set of contextually given alternatives to Mary’ (Rooth 1992). From the prosodic point of view, informational focus is normally associated with neutral stress, while contrastive focus is in general associated with strong or heavy stress (Chomsky and Halle 1968; Costa 1998: 200; or in Reinhart (1995) terms, “neutral stress” and “marked stress”). Sometimes contrastive focus and informational focus may coincide and, therefore, it is possible to find a phonological marking in the two types of focus.

Before analysing focus on possessives, it is important to emphasise that in most contexts containing possessives it is the whole DP that is focalised, as Cardinaletti (1998: 44) also notes for Italian. See the following example:

(31) O MEU LIVRO, não o dicionário, está esgotado. THE MY BOOK not the dictionary is out of print ‘My book, not the dictionary, is out of print.’

In (31), through the opposition between o meu livro and o dicionário, the two DPs are contrasted and, therefore, this behaviour does not help for the comprehension of the nature of the possessives, because the whole DP is subject to a heavy stress.

It is important, then, to study the intervention of possessives by themselves in focalisation in several contexts. A brief reference to vocatives, as illustrated in (32), where the possessive is isolated and cannot be preceded by an article:

(32) MEU senhor, ajudai-me! ‘My lord, help me!’

If vocatives, being referential, are full DPs (Coene and D’Hulst 2003a: 7), the possessive is in D and is intrinsically definite. So, independently of a more detailed analysis, the example (32) shows that an isolated pre-nominal possessive may be focalised and receive heavy stress. Let us now analyse focus on possessives in full DPs; before the analysis of EP, it is important to remember the relevant data in Italian.

4.3.2. Italian In Italian, the use of a pre-nominal possessive as informational focus, as in answers to wh questions, gives rise to ungrammaticality; the only way to answer to (33a) is by using a post-nominal possessive (33c) (Cardinaletti 1998: 21):

(33) a. La macchina di chi ti ha investito? the car of whom you has run over ‘Whose car has run over you?’ b. * la sua macchina the his/her car ‘his/her car’ c. la macchina sua the car of his/of her ‘his/her car’

Moreover, Cardinaletti (1998: 19) shows that in (34) the interpretations of la sua casa and la casa sua are slightly different and that a post-nominal possessive is used with a definite

Ana Maria Brito

36

determiner only if it is (contrastively) focalised, as in (34a); clearly, a pre-nominal possessive cannot be focalised in Italian, as in (34b):

(34) a. La casa SUA, non tua. the house his/her not your ‘His house, not yours.’ b. * La SUA casa, non tua. the his/her house not your

These data lead Cardinaletti to propose that in Italian post-nominal possessives are strong forms and pre-nominal possessives are weak forms. Nevertheless the author assumes that both pre-nominal and post-nominal possessives are XP.

4.3.3. European Portuguese Let us see now how EP possessives behave in terms of the marking of informational and contrastive focus in full DPs. Castro and Costa (2002: 104) had already shown that both pre-nominal and post-nominal possessives in EP may express informational focus, as illustrated in examples (35) and (36):

(35) a. Viste o carro de quem? saw the car of whom ‘Whose car did you see? b. o seu carro / o meu carro / o teu carro the his car the my car the your car ‘his car / my car / your car’

(36) a. Foste à Feira do Livro? Compraste um livro de quem? went to the Book Fair bought a book of whom ‘Did you go to the Book Fair? Whose book did you buy?’ b. um livro meu / um livro teu / um livro seu / dele ‘a book of mine / a book of yours / a book of him’

The examples show that both pre-nominal and post-nominal possessives, answering wh questions, may introduce a new referent in the discourse. EP differs in this respect from Italian. But a more detailed analysis on focus on possessives both in definite and indefinite contexts is necessary. Let us consider examples (37), with definites.

(37) a. O meu livro está esgotado. the my book is out of print ‘My book is out of print.’ b. O MEU livro está esgotado. ‘My book is out of print.’ c. O MEU livro é que está esgotado. the my book is that is out of print ‘It is my book that is out of print.’ d. ??O MEU livro, não teu, está esgotado. the my book not your is out of print ‘My book, not yours, is out of print.’

The examples show that a pre-nominal possessive with definite article may be focalised by a prosodic process (a heavy stress) (37a), by a cleft sentence (37b) or by a contrastive context (37c). The example (37d) is not accepted by all speakers, showing that, in a definite context, there is a preference for the contrast focalisation of the whole DP and not of the possessive itself; we will come back to this matter later on, when we analyse the indefinite contexts.


37

Castro and Costa (2002: 103-04 and 2003), and following them Costa and Szendröi (2006: 321), consider that “in Portuguese, unlike Italian, a focalised possessive is not post-nominal”, as in examples (38), (39) and (40):

(38) a. * o livro meu the book my ‘my book’ b. o MEU livro the my book ‘my book’

(39) a. O MEU problema é que não percebo nada disto. the my problem is that not understand.1SG nothing of.this ‘My problem is that I don’t understand a word of this.’ b. * O problema meu é que não percebo nada disto. the problem my is that not understand.1SG nothing of.this ‘My problem is that I don’t understand a word of this.’ (Castro and Costa 2002:104)

(40) a. Esse é o MEU problema, não o teu.6 this is the my problem not the your ‘This is my problem, not yours.’ b. * Esse é o problema meu, não o teu. this is the problem my not the your ‘This is my problem, not yours.’ (Castro and Costa 2002: 104)

Castro and Costa use these facts in order to propose that a pre-nominal possessive is X and not XP, contrary to what is proposed by Cardinaletti for Italian (p. 108). Also Costa and Szendröi (2006: 322), commenting on these facts, write: “The case of possessives in European Portuguese is an instance of stress-shift in a language that typically marks focus syntactically, as most Romance languages do.” In fact, it is very difficult to stress the final head N in a definite context, even if we want to focalise it; examples like (41) are marginal or even ungrammatical:

(41) a. ??O meu LIVRO está esgotado. the my book is out of print ‘My BOOK is out of print.’ b. *??O meu LIVRO é que está esgotado. the my book is that is out of print ‘It is my BOOK that is out of print.’ c. *??O meu LIVRO, não dicionário, está esgotado. the my book not dictionary is out of print ‘My BOOK, not dictionary, is out of print.’

Applying the classical algorithms of focus, what we would expect would be focus on the rightmost constituent, but this does not happen here. So, something must be said about the “stress-shift” on the possessive. I would like to propose that the focalisation of the pre-nominal possessive in a definite context is indeed a marked case and that the mechanism responsible for the “stress-shift” that applies on the possessive in a definite DP is contrastive focus assignment and not informational focus. This contrastive focalisation at the left side of the DP is not exclusive of possessives, because it can also affect quantifiers, as in (42):

(42) a. Eu encontrei TODOS os documentos (e não apenas alguns). ‘I found ALL the documents and not only some.’

6 Note that in this example, from Castro and Costa (2002, 2003), what seems to be focalized is the whole DP.

Ana Maria Brito

38

b. Eu encontrei TRÊS / ALGUNS / VÁRIOS documentos (e não todos). ‘I found THREE / SOME / SEVERAL documents, and not all.’

Frota (1992, 1994) argued that contrastive focus in Portuguese is a phonological category that is freely assigned. Being freely assigned, what seems plausible is that contrastive focus may affect not only X constituents but also XP constituents.

Coming back to focalised possessives in definite contexts, as in (38a) *o livro meu there are two reasons for its ungrammaticality: first, the focalisation of a post-nominal possessive, as argued by Castro and Costa (2002, 2003), and second, the co-occurrence of a definite article and a post-nominal possessive, which is not expected in Modern Portuguese. Note, however, that in a very formal register we still can find the post-nominal possessive with definite article in European Portuguese, as in the following words of the prayer “Our father”:

(43) O pão nosso de cada dia nos dai hoje. the our bread of every day give us today ‘Give us our bread of every day.’

In medieval texts we can find examples like (44):

(44) a calça sua (apud Mattos e Silva 1989: 174) the trousers his ‘your trousers’

So, what we have in (43) is an option that can be related not only to a very conservative and formal construction but also to informational structure: if a language has a dominant grammar of possessives with post-nominal possessives and indefinites and pre-nominal possessive with definites and allows the post-nominal position with definites, then this is the signal of a focalisation of the possessive at the rightmost position (cf. also Miguel 2002:220).

Note that the post-position of focalised possessives with definite article may also rarely happen in Brazilian Portuguese, as in (45) (Neves 1993: 169):

(45) A televisão nossa está se fazendo. the television our is se making ‘One is making our television’

So, in this respect, BP is like Medieval Portuguese and Italian, accepting a focalised possessive in post-nominal position in a definite context. Let us see now focus on post-nominal possessives in indefinite contexts:

(46) a. Um livro meu está esgotado. a book my is out of print ‘A book of mine is out of print.’ b. Um livro MEU está esgotado. a book MY is out of print ‘A book of MINE is out of print.’ c. Um livro MEU é que está esgotado. a book my is that is out of print ‘It is a book of mine that is out of print.’ d. Um livro MEU, não TEU, está esgotado. a book my not your is out of print ‘A book of mine, not yours, is out of print.’

As we have seen before, in the dominant grammar of EP, possessives in indefinite contexts are post-nominal. The examples (48) show that post-nominal possessives may be focalised, differently to what is assumed by Castro and Costa (2002: 104 and 2003) and Costa and Szendröi (2006: 321). In an unmarked context, as in (46a), we have informational focus, the possessive is post-nominal and we have a neutral stress on the possessive. This possibility is


39

also expected from the discursive point of view, because informational focus is in general related to indefiniteness. In (46b), (c) and (d) we have a heavy stress and a contrastive focus on the possessive, favoured in (c) and (d) by a syntactic context that permits it.

Note that the indefinite environment is much better than the definite environment for the contrast of the possessive, as shown by examples like (46d), in comparison with (37d), here repeated as (47):

(47) ?? O MEU livro, não teu, está esgotado. the my book not your is out of print ‘My book, not yours, is out of print.’

This difference of grammaticality is important; in fact, if the pre-nominal possessive is tendentially X and the post-nominal possessive is XP, then the difference between (46d) and (47) may be explained by the fact that the contrastive focus preferably affects XP constituents than X constituents, differently to what Castro and Costa (2003) proposed. The following data confirm this hypothesis. In the grammar of EP that accepts pre-nominal possessives with indefinites, contrastive focus on the possessive is possible:

(48) a. Uns MEUS amigos, não TEUS, vieram visitar-me. a (some) MY friends not YOUR came to visit me ‘My friends came to visit me, not yours.’ b. Uns MEUS livros, não TEUS, foram publicados. a (some) MY books not YOUR were published ‘My books, not yours, were published.’

These examples show two things: first, pre-nominal possessives may be contrastively focalised, independently of the type of the determiner with which they co-occur, definite or indefinite; second, under contrastive focalisation pre-nominal possessives in an indefinite context, as in (48), are much better than pre-nominal possessives in a definite context, as in (47). This is certainly due to the category of these possessives: in (47), a marginal example, the possessive is tendentially X, while in (48) the possessive is XP. This shows that contrastively focalised possessives are XP, whether they are pre-nominal or post-nominal.

Let us then return to the question of the existence of grammars of pre-nominal possessives in competition in EP. 5. More on grammars of possessives in competition

We said in the section 3 that there are at least two grammars of pre-nominal possessives in EP. The non-dominant grammar allows pre-nominal possessives with indefinites, the dominant grammar only accepts pre-nominal possessives with definites and post-nominal ones with indefinites. Other facts corroborate this variation between grammars. Some speakers accept the temporal / aspectual adverbs ainda ‘still’ and já ‘already’ between the definite article and the possessive, as in (49a) and (49b) (Brito 2001: 509):

(49) a. a ainda minha mulher the still my wife b. a já sua noiva the already his fiancée

And some speakers even accept intensifier adverbs between the definite article and the possessive; the example (50a) is given by Miguel (2002a: 294), showing that in this grammar possessives are degree words; the intensifier adverb can of course appear more easily in post-nominal position (50b):

Ana Maria Brito

40

(50) a. O muito meu e muito teu amigo de Paris chega amanhã. the very my and very your friend from Paris will arrive tomorrow b. Aqui tens uma obra muito minha. here you have a work very mine

So, the non-dominant grammar, essentially described by Brito (2001, 2003) and Miguel (2002 a, b, c and 2004), allows pre-nominal possessives with indefinites, allows some temporal / aspectual adverbial modification between the definite article and the possessive; for some speakers possessives have degree and accept intensifier adverbs and, therefore, there is no adjacency requirement between determiners and possessives. In this grammar, there is the possibility of coordination and focalisation of pre-nominal possessives, showing that pre-nominal possessives are XP (see also Miguel 2004: 402).

In the dominant grammar, essentially described by Castro and Costa (2002, 2003) and Castro (2005), there is an adjacency requirement between possessives and determiners, the pre-nominal possessives are used exclusively with definite articles and these words have no degree. These behaviours indicate that in this grammar pre-nominal possessives are tendentially heads, X. But it is possible to find in the dominant grammar cases of coordination of pre-nominal possessives associated to focalisation, what indicated that the speakers of this grammar produce some occurrences where pre-nominal possessives are XP.

Remember also that certain southern and oral dialects of EP exhibit reduced pre-nominal possessives. This means, as Miguel (2002c and 2004) has already noted, that there is a third possibility in EP, a non-dominant grammar where pre-nominal possessives are behaving as heads, X, almost clitics.

In next sections we will analyse syntactically these grammars in competition. 6. A syntactic analysis of possessives in EP

6.1. The categorial nature of possessives: a bit of history

Before the proposal of a syntactic analysis of the three grammars of possessives in EP, let us consider more carefully what a possessive is. Possessives must be considered in two ways: related to person pronouns and related to the Noun they co-occur with. Possessives express, in fact, a double relation: first, a relation to the three discourse persons—the 1st, the 2nd and the 3rd persons both singular and plural; from this point of view, possessives are person pronouns and have a clear deictic dimension; and second, they relate the persons of the discourse to the entities designated by the N. So, when related to person pronouns, possessives share the following features: [+ N], [+ person], [+ number] (related to grammatical person). Depending on the language, they may also have a [+ gender] feature (related to the third person).

Regarding semantic features: they are [+specific] (by default); in some languages, pre-nominal possessives express definite values (French, Spanish, English), in other ones, they are unspecified in relation to this feature. In EP, we have seen that there is a difference between grammars: for two grammars, possessives are [+definite], for another and more formal grammar they are [� definite].

Concerning the relation between possessives and Ns, possessives manifest genitive case and, in some languages, agreement: in EP this agreement manifests in number and gender (o meu amigo, a minha amiga, os meus amigos, as minhas amigas); in Spanish only in number (mi amigo, mi amiga, mis amigos, mis amigas); in French pre-nominal possessives vary in number but they manifest gender only in the singular (mon livre, ma maison, mes livres, mes maisons) (see also Castro 2005, chap. 2; and Miguel 2004, chap. 4).

In Classical Latin there were person pronouns in the genitive case (mei, tui, sui, nostri, vestri) and possessives, with an inherent genitive case and agreement (meus, tuus, suus, nostrus, vestrus); the two are illustrated in (51):


41

(51) a. pater meus b. pater mei

The examples in (51) mean the same thing (my father), but the form that was already a possessive adjective (meus) was more frequent than the genitive person pronoun (Ernout and Thomas 1989: 179-181).7 Classical Latin had no system of determiners and these ones would originate from different adjective forms and pronouns. With the birth of Romance languages and the emergence of weak personal pronouns and of articles in the first position of DP as a way to codify semantic and discourse notions, as topic, given, definite, non definite (Vincent 1997), what happens is that the order of possessives inside DP becomes freer than before. But possessives in Classical Latin were already inflected forms, as illustrated in (51).

In Medieval Portuguese, besides the pre and post-nominal positions already mentioned, there were weak forms, intrinsically definite (mha / ma, ta, sa) and strong forms (mia / minha, tua, sua), co-ocurring with articles (Mattos e Silva 1989: 174-185):

(52) a. toda sa sobérvia all his pride b. o seu homem the his man

So, the data of Medieval Portuguese indicate a tripartite system of possessives, as Miguel (2004: 389-93) already proposed.

What is important to emphasise now is that possessives, as for their morphological nature, their internal syntax, are essentially a collection of � features, some of them shared by person pronouns, and some of them associated to the relation between the possessive and the N, that is, they are �P, in terms of Déchaine and Wiltschko (2003). But they vary from language to language in their external syntax, basically in their determiner and adjective status, what will be considered in 5.2.

6.2. Adjectives or determiners? Giorgi and Longobardi’s Parameter

Starting with the different position occupied by pre-nominal possessives in languages like Italian, which accepts the co-occurrence between articles and pre-nominal possessives, and English and French, which do not accept this co-occurrence, Giorgi and Longobardi (1991: 154) proposed the following “Possessive Parameter”:

(53) Possessive elements are syntactically specified to be realized on the surface either as As (as in Italian) or as Ds (as in English and French).

At least for the group of languages we are analysing this proposal goes on the right track. But we know that there is a significant micro-variation in Romance languages in pre-nominal possessives with respect to the presence and the absence of the article; some examples are presented in (54), (55) and (56) (cf. Picallo 1994 and Cardinaletti 1998):

(54) a. il suo padre (standard Italian) the his father b. el me libro (Padova dialect) the my (reduced) book c. so pare de Toni (Padova dialect) his.CLITIC father of Toni (with doubling)

(55) a. mi casa (standard Spanish) my house 7 For an analysis of the evolution Latin-Romance Languages and, in particular, Old French and Old Portuguese

possessives see Miguel (2004, chap. 5).

Ana Maria Brito

42

b. la mi casa (Asturian) the my house

(56) a. el teu germà (Catalan) the your brother b. frade duo (Sardinian) brother his c. fratuta (Salentino) brother-your

Even in EP it is possible to have the two patterns, the isolated possessive being used with kinship nouns in a formal register (57b):

(57) a. A minha mãe saiu. the my mother left ‘My mother left.’ b. Minha mãe saiu. my mother left ‘My mother left.’

And in the dominant grammar of Brazilian Portuguese (BP) the absence of article before possessives is quite common (58a), although it is also possible to have the definite article followed by the possessive (58b):

(58) a. Minha carteira sumiu. my bag disappeared ‘My bag disappeared.’ b. A minha carteira sumiu. the my bag disappeared ‘My bag disappeared.’

These data show that pre-nominal possessives not only vary in their nature of D and A across languages but also in the same language, a variation that was already suggested in the presentation of grammars in competition of EP in the preceding sections. Note, however, that in the framework developed in the Minimalist Program, the variation between languages and the micro-variation within the same language is not due to different categorial status of words but to the features of lexical and functional categories. In this view, the variation and micro-variation in the possessive system is not only due to the possessives themselves but to the nature and features of D, in particular if D has some lexical material or not, how agreement (case, number and person) checking is done in each grammar, and, consequently, the different computational procedures that are instantiated (Chomky 2000).

6.3. Pre-nominal possessives in the DP structure in EP

With these reflections as a background, let us come back to the discussion about the DP structure containing pre-nominal possessives in the grammars in competition of EP we are analysing. Accepting the DP hypothesis and the idea that between DP and NP there are functional categories for agreement checking, in particular, Person and Number checking, I will propose that inside DP there are functional categories for checking �features, AgrP and NumP. But another functional category seems justified. As for the thematic nature of possessives and their relation to their position, it has been always remarked that possessives are not uniform in the semantic relationship they maintain with the N: some are part of the argument structure of the N, as internal or external arguments, as described in (59) (cf. Brito 1984, 2001; and Castro 2005: chap. 4):


43

(59) a. a suaAgent destruição ‘its/his/her (Agent) destruction’ b. a suaTheme destruição ‘its/his/her (Theme) destruction’

And other ones mean Possessor, which has been called in the literature an “R-relation”, not exactly a �-role (Higginbotham 1983; Grimshaw 1990; among others), as in (60):

(60) a sua casa the his/her house ‘his/her house’

In the first case, it can be justifiably assumed that possessives are merged and check their �-role with the N in NP and move to Spec of NumP and to Spec of AgrP; as N raises to Num and to Agr (Cinque 1994) it checks agreement with the possessive and the correct word order is obtained. In the second case, where possessives mean Possessor and do not represent true arguments of the N, it cannot be proposed that they are merged within NP; in this case, it seems justifiable to propose that it may occupy the Spec position of another functional category above NumP, that Parodi (1994) for Spanish called PossP, which is equivalent to (Genitive) CaseP. 8

In Modern Portuguese the two functional categories AgrP and PossP seem, at first sight, not justified, because doubling is rare; see (61) and (62), the two from literary discourse:

(61) Montaigne explica pelo seu modo dele a variedade dos seus livros. Montaigne explains by his manner of him… ‘Montaigne explains by his manner...’

(Machado de Assis, apud Cunha and Cintra 1984: 322)

(62) Comigo me desavim minha senhora de mim. with.me REFL was angry.1SG my Lady of me ‘With me I was angry, my Lady of me.’

(Maria Teresa Horta, “Minha senhora de mim”)

Doubling was relatively frequent in Medieval Portuguese (63):

(63) h�� seu rei deles a/one their king of.them ‘a king of theirs’

(apud Mattos e Silva 1989: 177)

In spite of the rarity of doubling in Modern Portuguese, there are reasons to propose the existence of PossP or (Genitive) Case: DP genitives, including the “prepositional possessives”, the classification Castro (2005) uses for dele, dela, etc. and the presence of the so-called “post-nominal” possessives, that I will analyse below.

As for the motivation for the movement of the possessive: in the non-dominant grammar of possessives, where pre-nominal possessives are not characterised with the [+definite] feature, and the pre-nominal possessives may occur with indefinites, the motivation for the movement is the attraction by the features of Person in AgrP and Number in NumP (Zribi-Hertz 1999: Brito 2001, 2003).

As for the categorial status of the pre-nominal possessive: we have seen that they are tendentially XP; as, in this grammar, pre-nominal possessives are distinct from the determiners, they are AP (associated to �P), being what Miguel (2002b) calls “personal adjectives”.

8 Schloorlemmer (1998) uses PossP with a goal similar to AgrP. Miguel (2004, chap. 3), also uses a Genitive Phrase

above NumP for the analysis of DP genitives, but this proposal is not related to the study of post-nominal possessives, that I will defend below.

Ana Maria Brito

44

The following structure represents the syntax of a DP containing pre-nominal possessives in the non-dominant grammar of EP: 9

(64) DP | D’ ei

D AgrP | ri os XP Agr’ | ri meusi Agr PossP | ri livrosj [ti] Poss’ ru Poss NumP | Num’ ru Num NP [tj] | N’ | N [tj]

Occupying the Spec of a high functional category, the possessive leaves available not only the D position for a determiner (definite article, demonstrative and even indefinites) but also other low positions for adjectives and for some sorts of quantifiers:

(65) a minha linda carteira the my beautiful bag ‘my beautiful bag’

(66) os meus outros/muitos livros the my other/many books ‘my other books/my many books’

Other orders are also possible, showing that possessives occupy other final positions besides Spec of AgrP and check in situ their person and number features, as in (67):10

(67) os outros meus livros the other my books

In the non-dominant grammar represented by pre-nominal possessives that may be reduced, the analysis must be different: possessives are always definite and exhibit clitic properties. Then, the motivation for the movement of possessives is the attraction of the possessive by the feature [+definite]; the relevant and simplified structure is the following:

9 For an overview of recent approaches of the DP structure see Coene and D’Hulst (2003a). 10 For a more detailed analysis see Miguel (2002a, b, c and 2004).


45

(68) DP | D’ ei D AgrP ru | D Poss Agr’ | | ru os mei livrosj ... u NP | N’ | N [tj]

What about the dominant grammar of pre-nominal possessives, essentially described by Castro and Costa? We have seen that in this grammar pre-nominal possessives are tendentially heads and are employed with the definite article. The question that must be discussed is if in this dominant grammar pre-nominal possessives already occupy the D position, in an adjunction to D (as I suggested for the reduced clitic possessives in (68)), or if they are still adjectives, in the Spec position of a functional category (as in (64)). In order to answer to this question, let us take the analysis made by Castro and Costa (2002, 2003) and Castro (2005) of the possibility of interpolation of temporal / aspectual adverbs like já and ainda between articles and possessives, as in (49), here repeated as (69):

(69) a. a ainda minha mulher the still my wife b. a já sua noiva the already his fiancée

The authors propose that these are weak adverbs and that, in these circumstances, they also adjoin to the articles in D; they also consider that in (69) the adverbs já and ainda are modifiers of the possessive and not of the DP. It seems to me that this could be true for já, because, as they note, when the possessive is absent, the examples are bad:

(70) a. *a já noiva do João the already fiancée of John b. ??a já noiva dele the already fiancée of him

But with ainda this apparent interdependence is not the same, because the following examples are good:

(71) a. a ainda noiva do João the still fiancée of John b. a ainda noiva dele the still fiancée of him c. o ainda Presidente the still President

This shows that ainda does not modify the possessive alone and then must appear in a position different from the one occupied by the possessive; in these circumstances, it is not obvious that in (69a), in the sequence ainda minha mulher, the article and the possessive are in the same D

Ana Maria Brito

46

position, as Castro (2005: 222) assumes. One hypothesis is that the adverb is in adjunction to the functional category that contains the possessive and that the possessive is still an adjective, in Spec of a functional category.

Then, what I have been classifying as the dominant grammar of possessives in EP is not a homogenous grammar, because it is a mixture of properties of the other two grammars. Like in the non-dominant grammar represented by (68), where pre-nominal possessives are in D, pre-nominal possessives are intrinsically definite and must appear adjacent to D. Like in the non-dominant grammar described in (64), the grammar represented by Castro and Costa accepts the interpolation of aspectual / temporal adverbs, the coordination and focalisation of pre-nominal possessives, showing that pre-nominal possessives are XP, specifically AP, in a position different from the one of the definite articles.

6.4. Post-nominal possessives in the DP structure in EP

We have seen that there are languages, like Spanish, that have phonologically strong forms of post-nominal possessives, as in (72):

(72) la casa suya the house her

Parodi (1994)11 proposes that the post-nominal possessive with the meaning of Possessor is merged and remain in situ in Spec of PossP, in a structure similar to (64), and are spelled-out in their strong form (suya). With the movement of the N to the head of Agr, the post-nominal order of the strong possessive is explained.

In EP there is no difference between phonologically weak and phonologically strong forms of possessives, but the explanation presented above may also be applied to this language, as already suggested in the preceding section. Before developing this hypothesis, it is important to remember the main contexts where post-nominal possessives may appear. Possessives may occur immediately after the noun, as in indefinite contexts illustrated several times in this paper, as in (73):

(73) uma casa minha a house my ‘a house of mine’

They can also appear at the rightmost position, after internal arguments expressed by PPs (74a) or after adjectives (74b) (Brito 1984):

(74) a. um livro de histórias teu a book of stories your ‘a book of stories of yours’ b. um livro de histórias antigo teu a book of stories old your ‘an old book of stories of yours’

There are also post-nominal possessives that represent a Theme selected by iconic nouns, as in (75):

(75) uma fotografia tuaTheme a picture your ‘a picture of yours’

And with the same type of Ns, even two possessives may co-occur, one at the left of the N and one at the right of the N, as in (76):

11 As for su casa, Parodi (1994) proposes that the pre-nominal possessive, after the checking of agreement with the

noun, cliticizes to D and is spelled-out as a weak form.


47

(78) Uma tua fotografia minha está estragada. a your picture my is spoiled ‘A picture of mine of yours is spoiled.’

It was already proposed (Brito 1984: 468) that, in general, the readings of this example satisfy the Possessivization Principle (Giorgi and Longobardi 1991: 66-71): Agent … Theme, Possessor … Theme. In these circumstances, one could assume that the final position of the Theme possessive minha in (76) would be the internal position within NP. But the readings Theme … Possessor / Agent are also marginally possible (Brito 1984: 468), which means that minha is not within NP. So, I make the hypothesis that the final position occupied by minha in (76), independent of its meanings, is Spec of PossP, in order to check its Genitive Case and Person features. With this hypothesis in mind, let us look at the data presented in (74). Suppose that livro de histórias may function, in this particular case, as a whole, almost as a compound; in these circumstances, this sequence may be dislocated as a whole to Agr. If the possessive occupies the Spec of PossP, its post-nominal position in (74a) is obtained; and if the adjective antigo occupies the Spec position of an intermediate category between AgrP and PossP, the example (74b) is also explained (cf. Cinque 1994).

So, unlike the traditional view, according to which post-nominal possessives are generated at the right side of the N, it was proposed that they occupy the Spec of a functional category, and their final post-nominal position is obtained by N movement to the left, regardless of their interpretation. Remember that post-nominal possessives may be modified by exclusion adverbs, are not reduced, have human reading, accept intensifier adverbs and may be easily coordinated and focalised. So they are XP, more specifically, AP and �P. 7. Synopsis of possessives in EP

The figure presented below contains a synopsis of possessives in EP:

Possessives in European Portuguese 1. Pre-nominal possessives

1.1. Dominant grammar (cf. Castro and Costa 2002, 2003; Castro 2005) [+def] �P ([+genitive case], [+ number], [+ gender]) A (as they codify [+def] they are becoming gradually D) position: Spec of AgrP no modification by exclusion adverbs no reduction [- degree] admit temporal / aspectual adverbs interpolation* accept focalization and coordination associated to focus* (* these properties indicate that they are AP, in the Spec of AgrP)

1.2. Non-dominant grammar (southern and oral dialects) [+def] �P ([+genitive case], [+ number], [+ gender], sometimes poor morphology) D (almost as clitic to a lexical D) position: adjunction to D reduction [- degree] no modification no coordination

1.3. Non-dominant grammar (formal) (cf. Brito 2001, 2003; Miguel 2002 a, b, c, 2004) [� def] �P ([+genitive case], [+ number], [+ gender]) AP position: Spec of AgrP no modification by exclusion adverbs no reduction [+ degree] (admit intensifier adverbs) admit temporal / aspectual adverbs interpolation accept focalization and coordination associated to focus

Ana Maria Brito

48

2. Post-nominal possessives - def

�P ([+genitive case], [+ number], [+ gender])

AP [+ degree] (admit intensifier adverbs)

modification by exclusion adverbs admit temporal / aspectual adverbs

no reduction coordination

focus

8. Conclusions

The research undertaken in this paper led us to the following conclusions:

1. We could confirm the existence of grammars in competition in EP mainly centred on pre-nominal possessives (confirming Miguel 2002a, b, c, 2004). There is a formal and non-dominant grammar where pre-nominal possessives are independent from definiteness, being able to co-occur with indefinites, where possessives have degree and that admits interpolation of adverbs between articles and the possessives; in this grammar pre-nominal possessives are tendentially XP; the XP (AP) nature was essentially proved from the possibility of focalisation and coordination of pre-nominal possessives associated to focus. There is also a non-dominant grammar, limited to southern and oral dialects of EP, mainly characterised by the reduction of pre-nominal possessives, where the latter are heads, almost clitics to D, are associated to definiteness and are strictly adjacent to the determiners. The dominant grammar of pre-nominal possessives mixes properties of the other two: pre-nominal possessives have some behaviours of heads, X, but simultaneously they have behaviours of maximal projections, XP, in particular, they may be focalised and coordinated; they are still A, but, as they codify the [+def ] feature, they are tendentially Ds. In these three grammars, possessives are essentially a collection of � features, they are �P, in the sense of Déchaine and Wiltschko (2003).

2. We have extended our study to post-nominal possessives, proposing an uniform analysis: they occupy the Spec of PossP, regardless of their meaning and the post-nominal position is obtained by N movement to the left.

3. The study of focus on possessives, both in definite and indefinite contexts, allowed to show that pre and post-nominal possessives may be focalized: informational focus on possessives is rightmost for prosodic convergence and it is, then, the unmarked case; as for contrastive focus, it explains the “stress-shift” on pre-nominal possessives not only in definite contexts but also in indefinite contexts in the non-dominant grammar of EP; contrastive focus is a phonological free category and may affect not only heads (X), Quantifiers and Numerals, but also maximal projections (XP), possessives. With respect to focus, EP is then distinct from Italian: in Italian, as already suggested by Castro and Costa (2003), focus assignment seems sensitive to the categorial nature: only XP forms can be focalized; in Portuguese focus assignment may affect X and XP constituents. The possibility of focus on pre-nominal and post-nominal possessives favors a framework without specific functional categories for informational structure in the DP, similar to clausal domain (Costa 1998: 191). So, in the DP domain, as well as in clausal domain, Syntax and Prosody play a role in the identification of focus in EP.


49

References

BRITO, A. 1984. Sobre as noções de sujeito e argumento externo: semelhanças e diferenças entre a estrutura de F e a estrutura de SN em Português. Boletim de Filologia XXIX: 421-478.

BRITO, A. 2001. Presença / ausência de artigo antes de possessivo no Português do Brasil. In Actas do XVI Encontro da Associação Portuguesa de Linguística, 551-575. Lisboa: APL/ Colibri.

BRITO, A. 2003. Os possessivos em Português Europeu numa perspectiva de Sintaxe comparada. Revista da Faculdade de Letras, Línguas e Literaturas [Universidade do Porto] II Série, Volume XX, Tomo II: 495-522.

CARDINALETTI, A. 1998. On the deficient / strong position in possessive systems. In A. ALEXIADOU and C. WILDER, eds. Possessors, Predicates and Movement in the Determiner Phrase, 17-53. Amsterdam: John Benjamins.

CARDINALETTI, A. 1999. Pronouns in Germanic and Romance languages. In H. van RIEMSDIJK, ed. Clitics in the Languages of Europe, 33-82. Berlin: Mouton de Gruyter.

CARDINALETTI, A. and M. STARKE. 1994. The typology of structural deficiency: On the three grammatical classes. In H. van RIEMSDIJK, ed. Clitics in the Languages of Europe, 145-233. Berlin: Mouton de Gruyter.

CASTRO, A. 2005. Possessives in European Portuguese. Ph.D., Universidade Nova de Lisboa and Paris VIII-Saint Denis.

CASTRO, A. and J. COSTA. 2002. Possessivos e advérbios: formas fracas como Xº. In Actas do XVII Encontro da Associação Portuguesa de Linguística, 101-111. Lisboa: APL / Colibri.

CASTRO, A. and J. COSTA. 2003. Weak forms as Xº: Prenominal possessives and preverbal adverbs in Portuguese. In A. T. PÉREZ-LEROUX and Y. ROBERGE, eds. Romance Linguistics: Theory and Acquisition, 95-110. Amsterdam: John Benjamins.

CHOMSKY, N. 2000. Minimalist Inquiries: The framework. In R. MARTIN, D. MICHAELS and J. URIAGEREKA, eds. Step by Step. Essays on Minimalist Syntax in Honor of Howard Lasnik, 89-155. Cambridge MA: MIT Press.

CHOMSKY, N. and M. HALLE. 1968. The sound pattern of English. Cambridge MA: MIT Press. CINQUE, G. 1993. A null theory of phrase and compound stress. Linguistic Inquiry 24.3: 239-297. CINQUE, G. 1994. On the evidence for partial N-movement in the Romance DP. In G. CINQUE et alii, eds.

Paths towards Universal Grammar. Studies in Honor of Richard Kayne, 85-110. Georgetown: Georgetown University Press.

COENE, M. and Y. D’HULST. 2003a. Introduction: The syntax and semantics of noun phrases. Theoretical background. In M. COENE and Y. D’HULST, eds. From NP to DP. Volume I: The Syntax and Semantics of Noun Phrases, 1-33. Amsterdam: John Benjamins.

COSTA, J. 1998. Word Order Variation. A Constraint-based Approach, Ph. Dissertation, Leiden University.

COSTA, J. and K. SZENDRÖI. 2006. Acquisition of focus marking in European Portuguese. Evidence for a unified approach to focus. In V. TORRENS and L. ESCOBAR, eds. The Acquisition of Syntax in Romance Languages, 319-329. Amsterdam: John Benjamins.

CUNHA, C. and L. F. L. CINTRA. 1984. Nova Gramática do Português Contemporâneo. Lisboa: Edições Sá da Costa.

ERNOUT, A. and F. THOMAS. 1989. Syntaxe Latine. Paris: Ed. Klincksieck. DÉCHAINE, R.-M. and M. WILTSCHKO. 2003. On pro-nouns and other “pronouns”. In M. COENE and Y.

D’HULST, eds. From NP to DP. Volume I: The Syntax and Semantics of Noun Phrases, 71-89. Amsterdam: John Benjamins.

FROTA, S. 1992. Para a Prosódia da Frase: Quantificador, Advérbio e Marcação Prosódica (Somente alguns tópicos em foco). MA Diss., Faculdade de Letras da Universidade de Lisboa.

FROTA, S. 1994. Is Focus a phonological category in Portuguese? In P. ACKEMA and M. SCHLOORLEMMER, eds. Proceedings of Console I, 69-86. The Hague: Holland Academic Graphics.

GIORGI, A. and G. LONGOBARDI. 1991. The Syntax of Noun Phrases. Cambridge: Cambridge University Press.

Ana Maria Brito

50

GRIMSHAW, J. 1990. Argument Structure. Cambridge MA: MIT Press. HIGGINBOTHAM, J. 1983. Logical form, binding and nominals. Linguistic Inquiry 14: 395-420. KAYNE, R. 1975. French Syntax. The Transformational Cycle. Cambridge MA: MIT Press. KAYNE, R. 1994. The Antisymmetry of Syntax. Cambridge MA: MIT Press. KISS, K. É. 1996. Focus Operator and Information Focus. Working Papers in the Theory of Grammar.

Hungarian Academy of Sciences. MATTOS E SILVA, R. V. 1989. Estruturas Trecentistas. Elementos para uma gramática do Português

Arcaico. Lisboa: Imprensa Nacional-Casa da Moeda. MIGUEL, M. 2002a. Para uma tipologia dos possessivos. In Actas do XVII Encontro da Associação

Portuguesa de Linguística, 287-299. Lisboa: APL / Colibri. MIGUEL, M. 2002b. O estatuto categorial dos possessivos: possessivos e adjectivos. In Actas do Encontro

Comemorativo dos 25 anos do Centro de Linguística da Universidade do Porto, 191-202. Porto: CLUP.

MIGUEL, M. 2002c. Possessive pronouns in European Portuguese and Old French. Journal of Portuguese Linguistics 2: 214-240.

MIGUEL, M. 2004. O Sintagma Nominal em Português Europeu. Posições de Sujeito, Ph. D. Diss., Universidade de Lisboa.

NEVES, H. M. 1993. Possessivos. In A. T. CASTILHO, ed. Gramática do Português Falado Vol. III: As Abordagens, 149-211. Campinas: FAPESP, Ed. da Unicamp.

PARODI, C. 1994. On case and agreement in Spanish and English DPs. In M. L. MAZZOLA, ed. Issues and Theory in Romance Linguistics. Selected Papers from the Linguistic Symposium on Romance Languages 1993. Washington DC: Georgetown UP.

PICALLO, M. C. 1994. Catalan possessive pronouns: The Avoid Pronoun Principle revisited. Natural Language and Linguistic Theory 12: 259-299.

POSTAL, P. 1969. On the so-called “Pronouns” in English. In D. REIDEL and S. SCHANE, eds. Modern Studies in English, 201-224. Englewood Cliffs NY: Prentice-Hall.

REINHART, T. 1995. Interface Strategies. OTS Working Papers in Linguistics, Utrecht University. RAPOSO, E. P. 1973. Sobre a forma o em Português. Boletim de Filologia 22: 361-415. SCHOORLEMMER, M. 1998. Possessors, articles and definiteness. In A. ALEXIADOU and C. WILDER, eds.

Possessors, Predicates and Movement in the Determiner Phrase, 55-86. Amsterdam: John Benjamins.

ROOTH, M. E. 1992. A theory of focus interpretation. Natural Language Semantics 1: 75-116. VINCENT, N.1997. The emergence of the D-system in Romance. In A. van KEMENADE and N. VINCENT,

eds. Parameters of Morphosyntactic Change, 149-169. Cambridge: Cambridge University Press. ZRIBI-HERTZ, A. 1999. Le système des possessifs en français standard moderne. Langue Française 122:

7-29. ZUBIZARRETA, M. L. 1998. Focus, Prosody and Word Order. Cambridge MA: The MIT Press.

Ana Maria Brito Faculdade de Letras da Universidade do Porto

Centro de Linguística da Universidade do Porto12 [email protected]

12 As a member of the Centre of Linguistics of the University of Porto (CLUP, Unidade I & D of FCT), the present

research was supported by FEDER / POCTI U0022/2003.

european portuguese possessives and the structure of dp

Documents